[Bug tree-optimization/58554] New: Revision 202619 causes runtime failure in CPU2006 benchmark 445.gobmk

[Bug tree-optimization/58554] New: Revision 202619 causes runtime failure in CPU2006 benchmark 445.gobmk

[pthaugen at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58554

            Bug ID: 58554
           Summary: Revision 202619 causes runtime failure in CPU2006
                    benchmark 445.gobmk
           Product: gcc
           Version: 4.9.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: pthaugen at gcc dot gnu.org
                CC: bergner at gcc dot gnu.org, dje.gcc at gmail dot com,
                    rguenth at gcc dot gnu.org
              Host: powerpc64-linux
            Target: powerpc64-linux
             Build: powerpc64-linux

gobmk started failing at runtime with the stated revision. Tracked down
offending code (from benchmark source engine/board.c) and reduced to the
following. Generated code is ignoring control dependence and simply calling
memset to set the entire array.

[pthaugen@igoo build_base_test_32.0000]$ cat junk.c
extern int board_size;
extern unsigned char board[421];

void clear_board(void)
{
  int k;

  for (k = 0; k < 421; k++) {
    /* Original:
    if (!((unsigned) (((k) / (19 + 1) - 1)) < (unsigned) board_size &&
(unsigned) (((k) % (19 + 1) - 1)) < (unsigned) board_size)) */
    if (k < board_size )
      board[k] = 3;
  }
}
[pthaugen@igoo build_base_test_32.0000]$
/home/pthaugen/install/gcc/trunk_work/bin/gcc -S -m32 -O3 junk.c

Generated assembler for the function:
clear_board:
    lis 3,board@ha
    li 4,3
    la 3,board@l(3)
    li 5,421
    b memset

[Bug tree-optimization/58554] [4.9 Regression] Revision 202619 causes runtime failure in CPU2006 benchmark 445.gobmk

[pinskia at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58554

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |wrong-code
   Target Milestone|---                         |4.9.0
            Summary|Revision 202619 causes      |[4.9 Regression] Revision
                   |runtime failure in CPU2006  |202619 causes runtime
                   |benchmark 445.gobmk         |failure in CPU2006
                   |                            |benchmark 445.gobmk
           Severity|normal                      |blocker

[Bug tree-optimization/58554] [4.9 Regression] Revision 202619 causes runtime failure in CPU2006 benchmark 445.gobmk

[law at redhat dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58554

Jeffrey A. Law <law at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
                 CC|                            |law at redhat dot com
         Resolution|---                         |DUPLICATE

--- Comment #1 from Jeffrey A. Law <law at redhat dot com> ---
Duplicate.

*** This bug has been marked as a duplicate of bug 58553 ***

[Bug tree-optimization/58554] [4.9 Regression] Revision 202619 causes runtime failure in CPU2006 benchmark 445.gobmk

[law at redhat dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58554

Jeffrey A. Law <law at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
   Last reconfirmed|                            |2013-09-27
         Resolution|DUPLICATE                   |---
     Ever confirmed|0                           |1

--- Comment #2 from Jeffrey A. Law <law at redhat dot com> ---
Since this doesn't depend on the recent threading changes to trigger, I'm
keeping this open as I'll probably revert a tiny piece of the threading changes
which will make 58553 go latent.

[Bug tree-optimization/58554] [4.9 Regression] Revision 202619 causes runtime failure in CPU2006 benchmark 445.gobmk

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58554

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot gnu.org

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
Mine.

[Bug tree-optimization/58554] [4.9 Regression] Revision 202619 causes runtime failure in CPU2006 benchmark 445.gobmk

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58554

--- Comment #4 from Richard Biener <rguenth at gcc dot gnu.org> ---
Ok, this one shows that pattern detection does not consider conditionally
executed stores properly.  I have a fix.

[Bug tree-optimization/58554] [4.9 Regression] Revision 202619 causes runtime failure in CPU2006 benchmark 445.gobmk

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58554

--- Comment #5 from Richard Biener <rguenth at gcc dot gnu.org> ---
Author: rguenth
Date: Mon Sep 30 12:22:00 2013
New Revision: 203031

URL: http://gcc.gnu.org/viewcvs?rev=203031&root=gcc&view=rev
Log:
2013-09-30  Richard Biener  <rguenther@suse.de>

    PR tree-optimization/58554
    * tree-loop-distribution.c (classify_partition): Require unconditionally
    executed stores for memcpy and memset recognition.
    (tree_loop_distribution): Calculate dominance info.

    * gcc.dg/torture/pr58554.c: New testcase.

Added:
    trunk/gcc/testsuite/gcc.dg/torture/pr58554.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/tree-loop-distribution.c

[Bug tree-optimization/58554] [4.9 Regression] Revision 202619 causes runtime failure in CPU2006 benchmark 445.gobmk

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58554

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|---                         |FIXED

--- Comment #6 from Richard Biener <rguenth at gcc dot gnu.org> ---
Fixed.

[Bug tree-optimization/58554] [4.9 Regression] Revision 202619 causes runtime failure in CPU2006 benchmark 445.gobmk

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58554

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |Ganesh.Gopalasubramanian@am
                   |                            |d.com

--- Comment #7 from Richard Biener <rguenth at gcc dot gnu.org> ---
*** Bug 58577 has been marked as a duplicate of this bug. ***